Adoptive transfer of antigen-specific T ceils has emerged as a promising approach for reconstituting immunity to viral pathogens and transmitting tumoricidal effector cells to tumor-bearing hosts. Recent advances in techniques for sensitizing and expanding functional cytotoxic T cells ex vivo which can survive and proliferate in vivo have increased the potential applications of this approach. Furthermore, recognition that normal and tumor-bearing human hosts can generate cytotoxic T cell responses to certain self proteins such as WT1, which are normally expressed only in fetal life but are differentially expressed by certain types of cancer, has suggested that these antigens may be particularly promising targets for cell-based immunotherapy. Our experimental work indicates that certain of these oncofetat proteins, specifically WT1, survivin and HER2/neu are differentially expressed by specific developmental tumors of childhood. Preliminary studies also indicate that T cells specific for immunogenic peptides from one of these proteins, WT1, can be regularly generated from normal individuals and a proportion of patients bearing WTI" tumors. These WT1 peptide specific T cells can recognize and lyse WTI* primary tumors expressing their restricting allele both in vitro and in vivo, in SCID mice bearing tumor xenografts. In this project, we propose to examine and compare the anti-tumor activity of T cells generated from patients or normal, HLA sharing individuals in response to immunogenic peptides of WT1 or survivin, focussing on the characteristics of the T cell responses generated in vitro and their tumor specific activity both in vitro and in vivo. Specifically we will: 1) to define the capacity of children with malignancies of developing tissues, including neuroblastoma, Ewing's sarcoma, osteogenic sarcoma, rhabdomyosarcoma and DSRCT to generate cytotoxic T cell responses specific for known HLA-binding peptide epitopes of specific proteins differentially expressed by their own tumors, either WT1 or survivin. 2) to comparatively assess novel approaches for generating tumor-peptide specific T cells from donors expressing HLA alleles other than HLA A0201, identify novel epitopes presented by other HLA alleles, and comparatively assess the in vitro activity of T cells specific for these new epitopes as they are expressed by tumors. 3) to comparatively assess the in vivo tumor-specific targeting and tumoricidal activity of in vitro generated, HLA-restricted T cells specific for peptide epitopes of WT1 or survivin and functional subsets thereof in SCID mice bearing human tumor xenografts and 4) to conduct phase 1/11clinical trials of peptide specific autologous and thereafter, HLA-matched allogeneic T cells for adoptive cell immunotherapy of patients with chemotherapy refractory pediatric solid tumors.